Generative AI / Agentic AI Engineer with 3+ years at Accenture, shipping production AI for Fortune 10/100 clients across healthcare, aviation, and entertainment.
- π€ Designed a multi-agent failure-triage system that cut resolution time from 1β2 days β under 10 minutes
- π₯ Architected a real-time healthcare call assistant processing 5,000+ live utterances/day with sub-second latency and 99.5% uptime
- π¬ Delivered a GPT-4o document automation suite for a Fortune 100 entertainment company β 70β80% effort reduction across 15+ departments
- π§ͺ Built 8 open-source GenAI projects with 2,400+ tests spanning the full GenAI stack
- π Based in Noida, India β open to GenAI / Agentic AI opportunities
| # | Project | What it does | Key Tech | Tests |
|---|---|---|---|---|
| 1 | DocuMind | Production-grade RAG β PDF ingestion, hybrid search, Cohere reranking, cited multi-LLM answers & RAGAS evaluation | RAG Β· BM25 Β· Reranking Β· FastAPI | 9 |
| 2 | AgentForge | Multi-agent AI framework with tool-using agents, ReAct reasoning, orchestration patterns & Streamlit dashboard | ReAct Β· Tool Use Β· LangGraph Β· CrewAI | 62 |
| 3 | ModelForge | LLM fine-tuning pipeline β QLoRA on Mistral 7B with BLEU/ROUGE/BERTScore eval, LLM-as-Judge & cost analysis | QLoRA Β· PEFT Β· Mistral 7B Β· HuggingFace | 577 |
| 4 | GuardianAI | LLM safety & observability β content filtering, PII detection, token/cost tracking & real-time dashboard | Guardrails Β· PII Β· Drift Β· Streamlit | 457 |
| 5 | ShipAI | ML deployment pipeline β Docker, FastAPI serving, 3-tier health checks, GitHub Actions CI/CD & load testing | Docker Β· FastAPI Β· CI/CD Β· Locust | 650 |
| 6 | VisionRAG | Multimodal RAG β PDF + image ingestion with OCR, vision embeddings, hybrid retrieval & multimodal generation | OCR Β· Vision Β· Multimodal Β· RAG | 293 |
| 7 | InterviewGPT | AI mock interview system β adaptive difficulty, 14 heuristic scoring functions & progress analytics | Adaptive Engine Β· Heuristics Β· OpenAI | 295 |
| 8 | ToolBridge | Prompt engineering toolkit β template management, chain-of-thought, few-shot optimization, A/B testing & versioning | Prompt Eng Β· A/B Testing Β· Versioning | 95 |
π Total: 2,438 tests across 8 production-grade projects
Sparkling Star Award
2024 Β· Accenture & Avanade Recognized for building a multi-agent GenAI system adopted across enterprise workflows
ACE Award (Team)
2023 Β· Accenture Awarded for the real-time healthcare AI assistant β 5,000+ daily utterances, sub-second latency, 99.5% uptime
GenAI Portfolio Β· 8 Projects Β· 2,400+ Tests
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β β
β π DocuMind π€ AgentForge 𧬠ModelForge β
β Production RAG Multi-Agent AI LLM Fine-Tuning β
β β
β π‘οΈ GuardianAI π ShipAI ποΈ VisionRAG β
β Safety & Observ. ML Deployment Multimodal RAG β
β β
β π― InterviewGPT π§ ToolBridge β
β Mock Interviews Prompt Engineering β
β β
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β²γγγγ
β
βββββββββββββββββ΄βββββββββββββββββ
β Live Portfolio Site (GH Pages) β
β aditya2425.github.io/Portfolio β
ββββββββββββββββββββββββββββββββββ
β If you find any of these useful, a star goes a long way.
Open to GenAI / Agentic AI roles Β· Reach out: aditya78332412@gmail.com